[improvement](be) Optimize IN filter pruning with native HybridSet operations

[mrhhsg]

What problem does this PR solve?

Issue Number: None

Related PR: None

Problem Summary:

IN-filter metadata pruning previously converted HybridSet values through generic literal and Field materialization paths. This duplicated the IN set for zonemap pruning and made Bloom pruning allocate owning Field objects for every candidate.

This PR:

• exposes typed HybridSet operations for owning min/max bounds, exact range membership, and native-value callback traversal for Bloom probes;
• limits each zonemap snapshot to two owning non-null min/max Field values and performs exact checks against the source typed set;
• probes file Bloom filters directly from native HybridSet values with early exit, without building a Field vector;
• maintains fast min/max and range summaries for tinyint/smallint bitset containers;
• keeps dictionary, ORC, and Parquet pruning behavior conservative for unsupported or mismatched types.

Empty and NULL-only IN sets, embedded-NUL strings, borrowed string storage, prepared min/max reuse across expression clones, and native zonemap, dictionary, and Bloom pruning are covered by focused tests.

Release note

None

Check List (For Author)

• Test

  • Regression test
    • ./run-regression-test.sh --conf output/local-regression/regression-conf-19700.groovy --run -d query_p0/expr_zonemap -s test_expr_zonemap_pruning (1/1 suite passed)
    • ./run-regression-test.sh --conf output/local-regression/regression-conf-19700.groovy --run -d query_p1/expr_zonemap -s test_expr_zonemap_pruning_p1 (1/1 suite passed)
    • ./run-regression-test.sh --conf output/local-regression/regression-conf-19700.groovy --run -d query_p0/runtime_filter -s set_operator_in_filter (1/1 suite passed)
  • Unit Test
    • ./run-be-ut.sh --run --filter='ExprZonemapFilterTest.*:HybridSetTest.IntegerMinMaxAndRangeLookup:HybridSetTest.SignedBitSetRangeLookup:HybridSetTest.StringRangeLookupPreservesEmbeddedNull:RuntimeFilterMergerTest.*:RuntimeFilterProducerTest.*:FileScannerV2Test.ConditionCacheDigestIncludesRuntimeFilterPayload:ParquetBloomFilterPruningTest.NativeUint32BloomUsesPhysicalInt32Hash:ParquetBloomFilterPruningTest.NativeRowGroupKeepsPresentUint32AboveInt32Max:NewOrcReaderTest.SargDirectIn*' -j 32 (67/67 tests passed)
  • Manual test
    • BUILD_TYPE=ASAN ./build.sh --be --fe --be-extension-ignore paimon-scanner,preload-extensions -j 32
    • ./build-support/clang-format.sh
    • ./build-support/check-format.sh
    • CLANG_TIDY_BINARY=/tmp/doris-clang-tidy-with-resource-dir ./build-support/run-clang-tidy.sh --base 6ddb9af90571d5ceb41d13caacd80ebbf753b913 --build-dir be/ut_build_ASAN
    • git diff --check 6ddb9af90571d5ceb41d13caacd80ebbf753b913...HEAD
    • Focused benchmark list/run for (GetMinMax|LegacyMaterialize|AnyMatchRaw) (32/32 cases completed; benchmarked code is unchanged by the final scope cleanup).
  • No need to test or manual test. Explain why:

• Behavior changed:

  • No.
  • Yes.

• Does this need documentation?

  • No.
  • Yes.

Check List (For Reviewer who merge this PR)

• Confirm the release note
• Confirm test cases
• Confirm document
• Add branch pick label

[hello-stephen]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?